Joined: Sep. 2006
|Quote (RupertG @ Jan. 03 2009,09:05)|
|GilDodgen hasn't actually replied to any of the comments in UD that point out his misunderstanding of how modelling works, has he? It's as if they just don't exist... perhaps they don't fit his model. Certainly, everything he's written subsequently confirms the analysis someone made earlier that he's completely missed the point - and, perhaps, belongs in that class of people who are unable to work with that sort of abstraction.|
In a later blog post, GilDodgen does try to explain away his gross error.
|GilDodgen: In my original post about mutating the CPU instruction set, the OS, etc., I was being somewhat sarcastic. Obviously, this would be silly, and I wouldn’t expect anyone to take such an experiment seriously. My point was that if mutations are genuinely random, we should expect that in a biological system (e.g., a cell) they would interfere with or modify all aspects of a cell’s basic functioning, which would affect the ability of the cell to survive and reproduce. If random mutations killed off a significantly large percentage of cells or made them sterile before they had a chance to reproduce, pass on their genetic information, and for natural selection to work its magic, the rest of the simulation would be rendered invalid.|
(We observe mutations. It's not a secret. And sometimes they are deleterious. I have no idea why he brings up extreme mutation rates, but that can certainly be modeled. Too high a mutation rate, such as in a highly radioactive environment, and reproduction slows or ceases. This has nothing to do with mutating the computer OS or causing errors in the CPU.)
|GilDodgen: One other obvious point: A simulation must accurately depict the system being modeled. The computational machinery and information content of biological systems is inherent in, and quintessentially critical to, the function of the system being modeled, and therefore cannot be excluded from the effects of mutations, without the simulation being rendered completely meaningless.|
Yes, of the "biological systems", but the electronic computer is not "inherent in, quintessentially critical to, the function of the system being modeled". The simulation is an abstraction which is *independent* of the substrate on which it is being modeled. Of course, the 'organic computer' must be subject to mutation, but not the electronic computer being used to model the 'organic computer'. And of course, that's exactly what an evolutionary algorithm does.
Later on, GilDodgen takes on Word Mutagenation, but confuses the dictionary lookup (the blackbox fitness function) with the evolutionary algorithm. Of course, I was blocked from responding on Uncommon Descent.